Account rule, not a trading strategy
EVAL_ACCT_MGMT_V1Account rule: risk size for the Kraken $10k evaluationFAIL
tested Oct 24, 2026 11 variants tried
What was tested
This is not a trading strategy. It asks: given the evaluation rules (target +10%, a static drawdown floor of $9,400, a 3% daily loss limit, no time limit), is there a better way to size risk than a fixed $100 per trade?
Method: 20,000-path Monte Carlo from the balance of $9,954.28, using the trade results of our MF4 BTC strategy since 2024-10 (D1: 137 trades, −0.071R average) and a zero-edge baseline with random direction (D3). A historical replay was run as a cross-check.
Pre-registered pass bar and result
The bar was written and hashed before the run. Results are on data the selection step never saw (out-of-sample).
| Criterion | Bar | Result | Verdict |
|---|---|---|---|
| Ultimate pass probability ≥ fixed-$100 rule + 5 pp, D1 (recent MF4 trades) | ≥ 30.6% | best 30.2% (R_SNAP) | FAIL |
| Same, D3 (zero-edge baseline) | ≥ 32.2% | best 31.4% (R_SNAP) | FAIL |
| Not worse than the $100 rule at 60 days by more than 2 pp | > −2 pp | not recorded | INFO |
Why it failed
Without a time limit, risk size mostly changes speed, not the odds. The chance of ever reaching the target is about 25–28% for $100 to $250 a trade, against about 34.6% for a fair, cost-free game. The fixed $100 rule scored 25.6% (D1) and 27.2% (D3).
The bolder rule R_SNAP came closest (30.2% / 31.4%) but missed the bar and added a real chance of breaching the floor (8.9% / 7.6%). Most paths that do not pass end locked out by the buffer rather than breached. Verdict: keep $100. Adding a second instrument only made it faster (25.5% vs 25.6%).
Variants tried
11 rules: fixed risk of $50, 100, 150, 200, 250 or 300; a risk that scales with room above the floor; a bold “one winning trade reaches the target” size; a version that shrinks risk to fit the daily room; and two that halve the gap between trades (a stand-in for a second instrument).
What this test cannot tell
The Monte Carlo resamples trades as if independent, but real losses cluster. The historical replay gave a lower figure for the $100 rule (16.7%, with 15.7% still unresolved). The MF4 sample is only 137 trades.
Rules lock
Rules file SHA-256:
d665798820dbf3ec726192491529ac28afa158ae0c54b8d71d23c68d4f5acf64
Locked:
“not recorded” means the number or time was not written down at the time, and we have not filled it in afterwards.
Submit a rule
Have a rule with numbers? Send it to us on X @ProjectAnchorHQ or Telegram t.me/ProjectAnchorHQ (market, entry, exit, stop). Free; we only need the written rule, never keys, accounts or funds. We publish the result pass or fail, and credit you only if you say yes.